About

Currently at SambaNova, I apply advanced multimodal models including Llama3.2-Vision, Stormer, Hume, Chatterbox, and Wan2.1 onto specialized AI hardware. Previously at Phantom AI developing 2D and 3D object detection systems for autonomous driving.

I graduated with my MS and PhD from UC Berkeley and a dual degree from Seoul National University.

Experience

Senior Principal Engineer, Sambanova Systems [05/24 - present]
Staff Deep Learning Engineer, Phantom AI [05/17 - 05/24]
Research Intern, Lawrence Livermore National Lab [12 - 16 summers]
Research Assistant/Teaching Assistant, UC Berkeley [08/11 - 05/17]

Papers (show full / show selected)

(* indicates equal contribution)

Design and Synthesis of Multigrain Nanocrystals via Geometric Misfit Strain

Myoung Hwan Oh, Min Gee Cho, Dong Young Chung, Inchul Park, Youngwook Paul Kwon, Colin Ophus, Dokyoon Kim, Min Gyu Kim, Beomgyun Jeong, X. Wendy Gu, Jinwoung Jo, Ji Mun Yoo, Jaeyoung Hong, Sara McMains, Kisuk Kang, Yung-Eun Sung, A. Paul Alivisatos, Taeghwan Hyeon

Nature (cover) 2020
RoarNet: A Robust 3D Object Detection based on RegiOn Approximation Refinement

Kiwoo Shin*, Youngwook Paul Kwon*, Masayoshi Tomizuka

IV 2019
Deep Traffic Light Detection for Self-driving Cars from a Large-scale Dataset

Jinkyu Kim, Hyunggi Cho, Myung Hwangbo, Jaehyung Choi, John Canny, Youngwook Paul Kwon

ITSC 2018
A Novel Trajectory Prediction of Traffic Participants for Autonomous Lane Change Assistance

Donghan Lee*, Youngwook Paul Kwon*, Jinkyu Kim, Jongsang Suh

AVEC 2018
Convolution Neural Network-based Lane Change Intention Prediction of Surrounding Vehicles for ACC

Donghan Lee*, Youngwook Paul Kwon*, Sara McMains, J. Karl Hedrick

ITSC 2017
Using Machine Learning to Support Concept Clustering in Design Teams

Chengwei Zhang, Youngwook Paul Kwon, Julia Kramer, Euiyoung Kim, Alice Merner Agogino

JMD 2017
Deep Learning for Design in Concept Clustering

Chengwei Zhang, Youngwook Paul Kwon, Julia Kramer, Euiyoung Kim, Alice Merner Agogino

IDETC/CIE 2017
Artificial Intensity Remapping: Learning Multimodal Image Descriptors without Multimodal Image Data

Youngwook Paul Kwon, Sara McMains

NIPS-W 2016
DUDE (DUality DEscriptor): A Robust Descriptor for Disparate Images using Line Segment Duality

Youngwook Paul Kwon, Hyojin Kim, Goran Konjevod, Sara McMains

ICIP 2016
Energy-efficient Vector Field Based Toolpaths for CNC Pocket Machining

Sushrut Pavanaskar, Sushrut Pande, Youngwook Paul Kwon, Zhongyin Hu, Alla Sheffer, Sara McMains

JMP 2015
An Automated Grading/Feedback System for 3-view Engineering Drawings using RANSAC

Youngwook Paul Kwon, Sara McMains

L@S 2015

Patent

Data Augmentation Using Computer Simulated Objects for Autonomous Control Systems

Youngwook Paul Kwon, Phantom AI Inc.

US11714424B2 2023
Lane Line Reconstruction Using Future Scene and Trajectory

Youngwook Paul Kwon, Phantom AI Inc.

US11670173B2 2023
Method of Predicting Trajectory of Vehicle

Yeonhwa Cha, Myung Hwangbo, Youngwook Paul Kwon, Phantom AI Inc.

US11210533B1 2021

Youngwook Paul Kwon

About

Experience

Senior Principal Engineer, Sambanova Systems [05/24 - present]

Staff Deep Learning Engineer, Phantom AI [05/17 - 05/24]

Research Intern, Lawrence Livermore National Lab [12 - 16 summers]

Research Assistant/Teaching Assistant, UC Berkeley [08/11 - 05/17]

Papers (show full / show selected)

Design and Synthesis of Multigrain Nanocrystals via Geometric Misfit Strain

RoarNet: A Robust 3D Object Detection based on RegiOn Approximation Refinement

Deep Traffic Light Detection for Self-driving Cars from a Large-scale Dataset

A Novel Trajectory Prediction of Traffic Participants for Autonomous Lane Change Assistance

Convolution Neural Network-based Lane Change Intention Prediction of Surrounding Vehicles for ACC

Using Machine Learning to Support Concept Clustering in Design Teams

Deep Learning for Design in Concept Clustering

Artificial Intensity Remapping: Learning Multimodal Image Descriptors without Multimodal Image Data

DUDE (DUality DEscriptor): A Robust Descriptor for Disparate Images using Line Segment Duality

Energy-efficient Vector Field Based Toolpaths for CNC Pocket Machining

An Automated Grading/Feedback System for 3-view Engineering Drawings using RANSAC

Patent

Data Augmentation Using Computer Simulated Objects for Autonomous Control Systems

Lane Line Reconstruction Using Future Scene and Trajectory

Method of Predicting Trajectory of Vehicle